identify_subject is a data descriptor function used to identify a subject identifier variable in the data. If no identify_subject entry is provided then patient IDs will be constructed based on their row ordering in the data by the various estimation routines.

identify_subject(name, group = NULL)

Arguments

name

The name of the variable in the input data containing the subject identifier.

group

If there are repeated clusters of observations per subject, group identifies each group of observations. This is needed for data structures where each group may have multiple observations.

Value

A object that can be used by specify_models to add subject ID information to the data description object.

Details

If you are using a study design where there can potentially be multiple pseudo trials for a given patient in the data, then the subject ID can be used to appropriately group trial entries together, e.g. for cluster-based bootstrapping; importantly, as a result, you should always use identify_subject at a subject level rather than a pseudo trial level.

When there are multiple per-patient pseudo trials in conjunction with interval data, then each pseudo trial's entries must appear contiguously in the data and be sorted by time.

Author

M. Alan Brookhart